kernel: Alignment trap

kernel: Alignment trap

[JH]

Hi,

I am running 4G LTE (USB protocol) and WiFi on IMX6 board, the kernel
is 4.19.75 LTE. It is not clear where the kernel alignment trap came
from, the only thing I could see is the alignment trap message
generated immediate after USB GSM communication or mwifiex_sdio. Where
that the alignment trap came from? Which parts of the program could
contribute the kernel alignment trap, kernel iteself, or USB GSM or
mwifiex_sdio? Appreciate clues how to fix it.

Nov 20 05:08:09 solar kernel: usb 1-1: GSM modem (1-port) converter
now attached to ttyUSB0
Nov 20 05:08:09 solar kernel: option 1-1:1.2: GSM modem (1-port)
converter detected
Nov 20 05:08:09 solar kernel: usb 1-1: GSM modem (1-port) converter
now attached to ttyUSB1
Nov 20 05:08:10 solar kernel: mwifiex_sdio mmc0:0001:1: info: trying
to associate to 'Solar Analytics Wifi' bssid 78:8a:20:49:4b:c5
Nov 20 05:08:10 solar kernel: mwifiex_sdio mmc0:0001:1: info:
associated to bssid 78:8a:20:49:4b:c5 successfully
Nov 20 05:08:10 solar kernel: IPv6: ADDRCONF(NETDEV_CHANGE): mlan0:
link becomes ready
Nov 20 05:23:13 solar kernel: mwifiex_sdio mmc0:0001:1: info:
successfully disconnected from 78:8a:20:49:4b:c5: reason code 3
Nov 20 05:23:13 solar kernel: IPv6: ADDRCONF(NETDEV_UP): mlan0: link
is not ready
Nov 20 05:23:18 solar kernel: Alignment trap: not handling instruction
e8532f00 at [<b6c802b6>]
Nov 20 05:23:18 solar kernel: Unhandled fault: alignment exception
(0x001) at 0x38626667
Nov 20 05:23:18 solar kernel: pgd = 34bdb7e7
Nov 20 05:23:18 solar kernel: [38626667] *pgd=00000000

Thank you.

Kind regards,

- jh

Re: kernel: Alignment trap

[Greg KH]

On Wed, Nov 20, 2019 at 09:28:14PM +1100, JH wrote:
> Hi,
> 
> I am running 4G LTE (USB protocol) and WiFi on IMX6 board, the kernel
> is 4.19.75 LTE. It is not clear where the kernel alignment trap came
> from, the only thing I could see is the alignment trap message
> generated immediate after USB GSM communication or mwifiex_sdio. Where
> that the alignment trap came from? Which parts of the program could
> contribute the kernel alignment trap, kernel iteself, or USB GSM or
> mwifiex_sdio? Appreciate clues how to fix it.
> 
> Nov 20 05:08:09 solar kernel: usb 1-1: GSM modem (1-port) converter
> now attached to ttyUSB0
> Nov 20 05:08:09 solar kernel: option 1-1:1.2: GSM modem (1-port)
> converter detected
> Nov 20 05:08:09 solar kernel: usb 1-1: GSM modem (1-port) converter
> now attached to ttyUSB1
> Nov 20 05:08:10 solar kernel: mwifiex_sdio mmc0:0001:1: info: trying
> to associate to 'Solar Analytics Wifi' bssid 78:8a:20:49:4b:c5
> Nov 20 05:08:10 solar kernel: mwifiex_sdio mmc0:0001:1: info:
> associated to bssid 78:8a:20:49:4b:c5 successfully
> Nov 20 05:08:10 solar kernel: IPv6: ADDRCONF(NETDEV_CHANGE): mlan0:
> link becomes ready
> Nov 20 05:23:13 solar kernel: mwifiex_sdio mmc0:0001:1: info:
> successfully disconnected from 78:8a:20:49:4b:c5: reason code 3
> Nov 20 05:23:13 solar kernel: IPv6: ADDRCONF(NETDEV_UP): mlan0: link
> is not ready
> Nov 20 05:23:18 solar kernel: Alignment trap: not handling instruction
> e8532f00 at [<b6c802b6>]
> Nov 20 05:23:18 solar kernel: Unhandled fault: alignment exception
> (0x001) at 0x38626667
> Nov 20 05:23:18 solar kernel: pgd = 34bdb7e7
> Nov 20 05:23:18 solar kernel: [38626667] *pgd=00000000

You should have a much longer tracedump after this, right?  Can you
provide that please?

And why do you think this was a USB issue?

thanks,

greg k-h

Re: kernel: Alignment trap

[JH]

Hi Greg,

On 11/21/19, Greg KH <greg@kroah.com> wrote:
> On Wed, Nov 20, 2019 at 09:28:14PM +1100, JH wrote:
>> Hi,
>>
>> I am running 4G LTE (USB protocol) and WiFi on IMX6 board, the kernel
>> is 4.19.75 LTE. It is not clear where the kernel alignment trap came
>> from, the only thing I could see is the alignment trap message
>> generated immediate after USB GSM communication or mwifiex_sdio. Where
>> that the alignment trap came from? Which parts of the program could
>> contribute the kernel alignment trap, kernel iteself, or USB GSM or
>> mwifiex_sdio? Appreciate clues how to fix it.
>>
>> Nov 20 05:08:09 solar kernel: usb 1-1: GSM modem (1-port) converter
>> now attached to ttyUSB0
>> Nov 20 05:08:09 solar kernel: option 1-1:1.2: GSM modem (1-port)
>> converter detected
>> Nov 20 05:08:09 solar kernel: usb 1-1: GSM modem (1-port) converter
>> now attached to ttyUSB1
>> Nov 20 05:08:10 solar kernel: mwifiex_sdio mmc0:0001:1: info: trying
>> to associate to 'Solar Analytics Wifi' bssid 78:8a:20:49:4b:c5
>> Nov 20 05:08:10 solar kernel: mwifiex_sdio mmc0:0001:1: info:
>> associated to bssid 78:8a:20:49:4b:c5 successfully
>> Nov 20 05:08:10 solar kernel: IPv6: ADDRCONF(NETDEV_CHANGE): mlan0:
>> link becomes ready
>> Nov 20 05:23:13 solar kernel: mwifiex_sdio mmc0:0001:1: info:
>> successfully disconnected from 78:8a:20:49:4b:c5: reason code 3
>> Nov 20 05:23:13 solar kernel: IPv6: ADDRCONF(NETDEV_UP): mlan0: link
>> is not ready
>> Nov 20 05:23:18 solar kernel: Alignment trap: not handling instruction
>> e8532f00 at [<b6c802b6>]
>> Nov 20 05:23:18 solar kernel: Unhandled fault: alignment exception
>> (0x001) at 0x38626667
>> Nov 20 05:23:18 solar kernel: pgd = 34bdb7e7
>> Nov 20 05:23:18 solar kernel: [38626667] *pgd=00000000
>
> You should have a much longer tracedump after this, right?  Can you
> provide that please?

That was the last statement from the kernel log journalctl -t kernel
-b at the time.

> And why do you think this was a USB issue?

No, I was not saying it was a USB issue, nor did I know if it is a
kernel issue or mwifiex_sdio issue or something else, I don't have the
knowledge to determine it, that was why I have to go this list for
help, I did try to connect to Linux network mailing list linux-net or
kernel mailing list lkml, but I must get wrong with the mailing list
address, I could not get response or subscription frin linux-net or
lkml, if you could let me know the correct kernel network mailing
address and kernel mailing list address, I'll ask the help there as
well.

Thanks Greg.

- jh

Re: kernel: Alignment trap

[Peter Chen]

On 19-11-21 07:08:32, JH wrote:
> Hi Greg,
> 
> On 11/21/19, Greg KH <greg@kroah.com> wrote:
> > On Wed, Nov 20, 2019 at 09:28:14PM +1100, JH wrote:
> >> Hi,
> >>
> >> I am running 4G LTE (USB protocol) and WiFi on IMX6 board, the kernel
> >> is 4.19.75 LTE. It is not clear where the kernel alignment trap came
> >> from, the only thing I could see is the alignment trap message
> >> generated immediate after USB GSM communication or mwifiex_sdio. Where
> >> that the alignment trap came from? Which parts of the program could
> >> contribute the kernel alignment trap, kernel iteself, or USB GSM or
> >> mwifiex_sdio? Appreciate clues how to fix it.
> >>
> >> Nov 20 05:08:09 solar kernel: usb 1-1: GSM modem (1-port) converter
> >> now attached to ttyUSB0
> >> Nov 20 05:08:09 solar kernel: option 1-1:1.2: GSM modem (1-port)
> >> converter detected
> >> Nov 20 05:08:09 solar kernel: usb 1-1: GSM modem (1-port) converter
> >> now attached to ttyUSB1
> >> Nov 20 05:08:10 solar kernel: mwifiex_sdio mmc0:0001:1: info: trying
> >> to associate to 'Solar Analytics Wifi' bssid 78:8a:20:49:4b:c5
> >> Nov 20 05:08:10 solar kernel: mwifiex_sdio mmc0:0001:1: info:
> >> associated to bssid 78:8a:20:49:4b:c5 successfully
> >> Nov 20 05:08:10 solar kernel: IPv6: ADDRCONF(NETDEV_CHANGE): mlan0:
> >> link becomes ready
> >> Nov 20 05:23:13 solar kernel: mwifiex_sdio mmc0:0001:1: info:
> >> successfully disconnected from 78:8a:20:49:4b:c5: reason code 3
> >> Nov 20 05:23:13 solar kernel: IPv6: ADDRCONF(NETDEV_UP): mlan0: link
> >> is not ready
> >> Nov 20 05:23:18 solar kernel: Alignment trap: not handling instruction
> >> e8532f00 at [<b6c802b6>]
> >> Nov 20 05:23:18 solar kernel: Unhandled fault: alignment exception
> >> (0x001) at 0x38626667
> >> Nov 20 05:23:18 solar kernel: pgd = 34bdb7e7
> >> Nov 20 05:23:18 solar kernel: [38626667] *pgd=00000000
> >
> > You should have a much longer tracedump after this, right?  Can you
> > provide that please?
> 
> That was the last statement from the kernel log journalctl -t kernel
> -b at the time.
> 
> > And why do you think this was a USB issue?
> 
> No, I was not saying it was a USB issue, nor did I know if it is a
> kernel issue or mwifiex_sdio issue or something else, I don't have the
> knowledge to determine it, that was why I have to go this list for
> help, I did try to connect to Linux network mailing list linux-net or
> kernel mailing list lkml, but I must get wrong with the mailing list
> address, I could not get response or subscription frin linux-net or
> lkml, if you could let me know the correct kernel network mailing
> address and kernel mailing list address, I'll ask the help there as
> well.
> 

Do you have correct firmware has loaded for your device?
Do you have issues if building these two modules in kernel?

-- 

Thanks,
Peter Chen

Re: kernel: Alignment trap

[JH]

Hi Peter,

On 11/21/19, Peter Chen <peter.chen@nxp.com> wrote:
> On 19-11-21 07:08:32, JH wrote:
>> Hi Greg,
>>
>> On 11/21/19, Greg KH <greg@kroah.com> wrote:
>> > On Wed, Nov 20, 2019 at 09:28:14PM +1100, JH wrote:
>> >> Hi,
>> >>
>> >> I am running 4G LTE (USB protocol) and WiFi on IMX6 board, the kernel
>> >> is 4.19.75 LTE. It is not clear where the kernel alignment trap came
>> >> from, the only thing I could see is the alignment trap message
>> >> generated immediate after USB GSM communication or mwifiex_sdio. Where
>> >> that the alignment trap came from? Which parts of the program could
>> >> contribute the kernel alignment trap, kernel iteself, or USB GSM or
>> >> mwifiex_sdio? Appreciate clues how to fix it.
>> >>
>> >> Nov 20 05:08:09 solar kernel: usb 1-1: GSM modem (1-port) converter
>> >> now attached to ttyUSB0
>> >> Nov 20 05:08:09 solar kernel: option 1-1:1.2: GSM modem (1-port)
>> >> converter detected
>> >> Nov 20 05:08:09 solar kernel: usb 1-1: GSM modem (1-port) converter
>> >> now attached to ttyUSB1
>> >> Nov 20 05:08:10 solar kernel: mwifiex_sdio mmc0:0001:1: info: trying
>> >> to associate to 'Solar Analytics Wifi' bssid 78:8a:20:49:4b:c5
>> >> Nov 20 05:08:10 solar kernel: mwifiex_sdio mmc0:0001:1: info:
>> >> associated to bssid 78:8a:20:49:4b:c5 successfully
>> >> Nov 20 05:08:10 solar kernel: IPv6: ADDRCONF(NETDEV_CHANGE): mlan0:
>> >> link becomes ready
>> >> Nov 20 05:23:13 solar kernel: mwifiex_sdio mmc0:0001:1: info:
>> >> successfully disconnected from 78:8a:20:49:4b:c5: reason code 3
>> >> Nov 20 05:23:13 solar kernel: IPv6: ADDRCONF(NETDEV_UP): mlan0: link
>> >> is not ready
>> >> Nov 20 05:23:18 solar kernel: Alignment trap: not handling instruction
>> >> e8532f00 at [<b6c802b6>]
>> >> Nov 20 05:23:18 solar kernel: Unhandled fault: alignment exception
>> >> (0x001) at 0x38626667
>> >> Nov 20 05:23:18 solar kernel: pgd = 34bdb7e7
>> >> Nov 20 05:23:18 solar kernel: [38626667] *pgd=00000000
>> >
>> > You should have a much longer tracedump after this, right?  Can you
>> > provide that please?
>>
>> That was the last statement from the kernel log journalctl -t kernel
>> -b at the time.
>>
>> > And why do you think this was a USB issue?
>>
>> No, I was not saying it was a USB issue, nor did I know if it is a
>> kernel issue or mwifiex_sdio issue or something else, I don't have the
>> knowledge to determine it, that was why I have to go this list for
>> help, I did try to connect to Linux network mailing list linux-net or
>> kernel mailing list lkml, but I must get wrong with the mailing list
>> address, I could not get response or subscription frin linux-net or
>> lkml, if you could let me know the correct kernel network mailing
>> address and kernel mailing list address, I'll ask the help there as
>> well.
>>
>
> Do you have correct firmware has loaded for your device?

Not clear what the firmware you were talking about? I used open source
linux-firmware-sd8887, uBlox SARA-R4 and uBlox EMMY on IMX6. Let's say
we have been running this device for about 7 months, we used Kernel
5.1.0 before, I could not recall any error about the alignment trap in
kernel. We now moved to kernel 4.19.75 LTS, that is what it occurred
often, that worries me.

> Do you have issues if building these two modules in kernel?

No. I built the kernel and image from Yocto, there were no building issues.

Thank you Peter.

- jh

Re: kernel: Alignment trap

[Greg KH]

On Thu, Nov 21, 2019 at 05:06:45PM +1100, JH wrote:
> Hi Peter,
> 
> On 11/21/19, Peter Chen <peter.chen@nxp.com> wrote:
> > On 19-11-21 07:08:32, JH wrote:
> >> Hi Greg,
> >>
> >> On 11/21/19, Greg KH <greg@kroah.com> wrote:
> >> > On Wed, Nov 20, 2019 at 09:28:14PM +1100, JH wrote:
> >> >> Hi,
> >> >>
> >> >> I am running 4G LTE (USB protocol) and WiFi on IMX6 board, the kernel
> >> >> is 4.19.75 LTE. It is not clear where the kernel alignment trap came
> >> >> from, the only thing I could see is the alignment trap message
> >> >> generated immediate after USB GSM communication or mwifiex_sdio. Where
> >> >> that the alignment trap came from? Which parts of the program could
> >> >> contribute the kernel alignment trap, kernel iteself, or USB GSM or
> >> >> mwifiex_sdio? Appreciate clues how to fix it.
> >> >>
> >> >> Nov 20 05:08:09 solar kernel: usb 1-1: GSM modem (1-port) converter
> >> >> now attached to ttyUSB0
> >> >> Nov 20 05:08:09 solar kernel: option 1-1:1.2: GSM modem (1-port)
> >> >> converter detected
> >> >> Nov 20 05:08:09 solar kernel: usb 1-1: GSM modem (1-port) converter
> >> >> now attached to ttyUSB1
> >> >> Nov 20 05:08:10 solar kernel: mwifiex_sdio mmc0:0001:1: info: trying
> >> >> to associate to 'Solar Analytics Wifi' bssid 78:8a:20:49:4b:c5
> >> >> Nov 20 05:08:10 solar kernel: mwifiex_sdio mmc0:0001:1: info:
> >> >> associated to bssid 78:8a:20:49:4b:c5 successfully
> >> >> Nov 20 05:08:10 solar kernel: IPv6: ADDRCONF(NETDEV_CHANGE): mlan0:
> >> >> link becomes ready
> >> >> Nov 20 05:23:13 solar kernel: mwifiex_sdio mmc0:0001:1: info:
> >> >> successfully disconnected from 78:8a:20:49:4b:c5: reason code 3
> >> >> Nov 20 05:23:13 solar kernel: IPv6: ADDRCONF(NETDEV_UP): mlan0: link
> >> >> is not ready
> >> >> Nov 20 05:23:18 solar kernel: Alignment trap: not handling instruction
> >> >> e8532f00 at [<b6c802b6>]
> >> >> Nov 20 05:23:18 solar kernel: Unhandled fault: alignment exception
> >> >> (0x001) at 0x38626667
> >> >> Nov 20 05:23:18 solar kernel: pgd = 34bdb7e7
> >> >> Nov 20 05:23:18 solar kernel: [38626667] *pgd=00000000
> >> >
> >> > You should have a much longer tracedump after this, right?  Can you
> >> > provide that please?
> >>
> >> That was the last statement from the kernel log journalctl -t kernel
> >> -b at the time.
> >>
> >> > And why do you think this was a USB issue?
> >>
> >> No, I was not saying it was a USB issue, nor did I know if it is a
> >> kernel issue or mwifiex_sdio issue or something else, I don't have the
> >> knowledge to determine it, that was why I have to go this list for
> >> help, I did try to connect to Linux network mailing list linux-net or
> >> kernel mailing list lkml, but I must get wrong with the mailing list
> >> address, I could not get response or subscription frin linux-net or
> >> lkml, if you could let me know the correct kernel network mailing
> >> address and kernel mailing list address, I'll ask the help there as
> >> well.
> >>
> >
> > Do you have correct firmware has loaded for your device?
> 
> Not clear what the firmware you were talking about? I used open source
> linux-firmware-sd8887, uBlox SARA-R4 and uBlox EMMY on IMX6. Let's say
> we have been running this device for about 7 months, we used Kernel
> 5.1.0 before, I could not recall any error about the alignment trap in
> kernel. We now moved to kernel 4.19.75 LTS, that is what it occurred
> often, that worries me.

Why did you move to an older kernel?  If it all works on a newer one,
please stick with that, or use 'git bisect' to find the commit that
breaks things.

thanks,

greg k-h

Re: kernel: Alignment trap

[JH]

Hi Greg,

On 11/21/19, Greg KH <greg@kroah.com> wrote:
> On Thu, Nov 21, 2019 at 05:06:45PM +1100, JH wrote:
>> Hi Peter,
>>
>> On 11/21/19, Peter Chen <peter.chen@nxp.com> wrote:
>> > On 19-11-21 07:08:32, JH wrote:
>> >> Hi Greg,
>> >>
>> >> On 11/21/19, Greg KH <greg@kroah.com> wrote:
>> >> > On Wed, Nov 20, 2019 at 09:28:14PM +1100, JH wrote:
>> >> >> Hi,
>> >> >>
>> >> >> I am running 4G LTE (USB protocol) and WiFi on IMX6 board, the
>> >> >> kernel
>> >> >> is 4.19.75 LTE. It is not clear where the kernel alignment trap
>> >> >> came
>> >> >> from, the only thing I could see is the alignment trap message
>> >> >> generated immediate after USB GSM communication or mwifiex_sdio.
>> >> >> Where
>> >> >> that the alignment trap came from? Which parts of the program could
>> >> >> contribute the kernel alignment trap, kernel iteself, or USB GSM or
>> >> >> mwifiex_sdio? Appreciate clues how to fix it.
>> >> >>
>> >> >> Nov 20 05:08:09 solar kernel: usb 1-1: GSM modem (1-port) converter
>> >> >> now attached to ttyUSB0
>> >> >> Nov 20 05:08:09 solar kernel: option 1-1:1.2: GSM modem (1-port)
>> >> >> converter detected
>> >> >> Nov 20 05:08:09 solar kernel: usb 1-1: GSM modem (1-port) converter
>> >> >> now attached to ttyUSB1
>> >> >> Nov 20 05:08:10 solar kernel: mwifiex_sdio mmc0:0001:1: info:
>> >> >> trying
>> >> >> to associate to 'Solar Analytics Wifi' bssid 78:8a:20:49:4b:c5
>> >> >> Nov 20 05:08:10 solar kernel: mwifiex_sdio mmc0:0001:1: info:
>> >> >> associated to bssid 78:8a:20:49:4b:c5 successfully
>> >> >> Nov 20 05:08:10 solar kernel: IPv6: ADDRCONF(NETDEV_CHANGE): mlan0:
>> >> >> link becomes ready
>> >> >> Nov 20 05:23:13 solar kernel: mwifiex_sdio mmc0:0001:1: info:
>> >> >> successfully disconnected from 78:8a:20:49:4b:c5: reason code 3
>> >> >> Nov 20 05:23:13 solar kernel: IPv6: ADDRCONF(NETDEV_UP): mlan0:
>> >> >> link
>> >> >> is not ready
>> >> >> Nov 20 05:23:18 solar kernel: Alignment trap: not handling
>> >> >> instruction
>> >> >> e8532f00 at [<b6c802b6>]
>> >> >> Nov 20 05:23:18 solar kernel: Unhandled fault: alignment exception
>> >> >> (0x001) at 0x38626667
>> >> >> Nov 20 05:23:18 solar kernel: pgd = 34bdb7e7
>> >> >> Nov 20 05:23:18 solar kernel: [38626667] *pgd=00000000
>> >> >
>> >> > You should have a much longer tracedump after this, right?  Can you
>> >> > provide that please?
>> >>
>> >> That was the last statement from the kernel log journalctl -t kernel
>> >> -b at the time.
>> >>
>> >> > And why do you think this was a USB issue?
>> >>
>> >> No, I was not saying it was a USB issue, nor did I know if it is a
>> >> kernel issue or mwifiex_sdio issue or something else, I don't have the
>> >> knowledge to determine it, that was why I have to go this list for
>> >> help, I did try to connect to Linux network mailing list linux-net or
>> >> kernel mailing list lkml, but I must get wrong with the mailing list
>> >> address, I could not get response or subscription frin linux-net or
>> >> lkml, if you could let me know the correct kernel network mailing
>> >> address and kernel mailing list address, I'll ask the help there as
>> >> well.
>> >>
>> >
>> > Do you have correct firmware has loaded for your device?
>>
>> Not clear what the firmware you were talking about? I used open source
>> linux-firmware-sd8887, uBlox SARA-R4 and uBlox EMMY on IMX6. Let's say
>> we have been running this device for about 7 months, we used Kernel
>> 5.1.0 before, I could not recall any error about the alignment trap in
>> kernel. We now moved to kernel 4.19.75 LTS, that is what it occurred
>> often, that worries me.
>
> Why did you move to an older kernel?  If it all works on a newer one,
> please stick with that, or use 'git bisect' to find the commit that
> breaks things.

My CTO wanted LTS and could not wait for 5.4, I did forward your blog
Linux Kernel Release Model for my CTO to read it, if it is caused by
kernel 4.19, I'll talk to the CTO again.

Thank you.

- jh